|
Post by cye on Aug 5, 2014 4:33:34 GMT -5
and here's another one i may have messed up... the black paint that i chose for my radiator was "barbecue paint".. i picked this so it would tolerate the high temperatures, but em, i'm now wondering (reading a bit), as bbq paint is "heat resistant" paint... could that be affecting the heat transfer? i don't think that's an issue with the paint and you shouldn't worry. paint is never a good insulator due to the fact that it is so thin. what they mean by heat resistant is that it will not peel off. i'd say it is exactly what you need, provided its not a very shiny black. cast iron stove paint is also good and will be something similar. the innards of that type of radiator are not complex, it's essentially two sheets (well one sheet folded in half, and with the corrugations missing at the fold) of corrugated heavy steel sheet, folded over on itself. The point of contact, i.e., the vertical narrowest points, are spot welded together. if you look carefully you should be able to see the spot welds. there are essentially then two large bore horizontal channels (forming header and footer 'manifolds') connected by a whole array of narrower bore vertical channels or 'risers'. Domestic radiators are available with ports both top and bottom, but unfortunately not all. a port cut off a scrap radiator could easily be brazed onto the flat outward face at the bottom left of that radiator, and whilst the average plumber would not have the skills, your local old-timer type motor mechanic may do. best regards cye
|
|
|
Post by cravings on Aug 5, 2014 5:21:19 GMT -5
well from a bit of reading around, it seems that only a very thick coating of any paint would make a considerable difference to the heat conductivity of a metal. some interesting reading from people who sound like they know a bit here for example. i sanded off the old white paint and there's only a very thin layer of black, so that's probably not a thing here. hopefully.
|
|
|
Post by cye on Aug 5, 2014 7:45:18 GMT -5
yes, paint thickness on that scale is nothing to worry about. never has anyone heard of a central heating system not working because there was an extra layer or two of paint on the rads. please let us know how you get along
|
|
|
Post by cravings on Aug 5, 2014 10:58:51 GMT -5
ok window was off, and everything inside the box is black. next move is to drain (again) and redo some of the joints, and change the nrv for a new one, fill again and hope for the best. if all watertight, then the lagging process begins.
next after that will be looking at a tilt for the panel, but i don't think i'm going to get much of a tilt there.
hot water can rise through a radiator in normal use, with the connections at the bottom.. water in and out on the same level..
|
|
|
Post by cye on Aug 5, 2014 12:43:53 GMT -5
ok window was off, and everything inside the box is black. next move is to drain (again) and redo some of the joints, and change the nrv for a new one, fill again and hope for the best. if all watertight, then the lagging process begins. next after that will be looking at a tilt for the panel, but i don't think i'm going to get much of a tilt there. hot water can rise through a radiator in normal use, with the connections at the bottom.. water in and out on the same level.. It is interesting to compare the workings of a radiator to that of a solar collector panel, it certainly took me back to first principles. When a rad is working as a standard radiator, heat rises inside the rad and the bulk of the rad is cooler than the incoming central heating water, so the hot water rises in the rad and this sets up a vertical up & down flow in the risers of the rad caused by thermosyphon as heat is then lost to the surrounding air by convection, & there is a continual upward flow of warming air rising up along the front and back surfaces of the radiator. This is why a curtain can often be seen to move when the rad below is on. To a much lesser extent a rad loses heat by by radiation , i.e., the name radiator is a misnomer. Now, with a flat plate solar panel, there is very little of the air convection thing going on. The panel captures most of its energy from radiation warming the black surface, and to a much lesser extent it will capture heat from the surrounding air. Insulating and glazing the box is as much about minimising convection losses as it is about gaining heat by convection. The most efficient way to capture heat off the black metal is to have a suitable minimum rate of flow of cooler liquid chilling it. you get less of that necessary flow from having the bulk of the flow rushing straight across the top of the panel as you are relying largely on the vertical thermosyphon current 'eddies' to cool the panel, and thermosyphon unfortunately provides a slower flow rate than is really required for efficient energy capture. it is therefore not quite the same as a central heating radiator in reverse but there are certainly similarities. the standard domestic radiator design is not very efficient at losing heat, and there are much more efficient designs available, but the standard design is cheap as chips and therefore commonplace. Even the standard rads would be more efficient at losing heat if the hot came in at the top and exited at the bottom, and indeed many classic rads have the ports for this, but it makes for more effort and materials to plumb and is not popular because it is untidier into the bargain. Yes, you will capture some of the heat with the current arrangement, no question about that, but not as much as can be captured by having a 'sufficient' flow of water across as much of the panel surface as possible. The work done by your pump with the current setup is largely being wasted as it is not delivering the required flow to the hot collector surface needing to be cooled. This is why all commercial flat panels use either a serpentine/snaked single pipe (strapped to a metal collector plate) which forces water to travel across all parts of the panel, or a header/footer/riser arrangement of pipes (strapped to a metal collector plate) with the cold entering at the bottom and the hot exiting at the top. Only in an evacuated tube solar panel does the water flow horizontally, but that technology is a different kettle of fish altogether as the heat pipes inside the vacuum tubes have a state change, liquid-to-gas & gas-to-liquid, thing going on. The more tilt you can get into that rad the more efficiently it will work as a collector panel as you will then get a greater flow of water across more of the collector surface. Keep up the good work. Hopefully you will see some satisfactory results.
|
|
|
Post by cravings on Aug 5, 2014 14:47:51 GMT -5
thanks again for all the help you're putting into the replies. much appreciated. so as said, got the innards of the panel sprayed black today. also sealed it up with tec7. drained the lot and swapped the NRV for a new one, and did indeed fix up the rest.. i've filled it up and had it pumping for a couple of hours and i can't see any more slow dripping, so i think it's finally done. next step is to lag the crap out of the whole lot, cylinder and all piping. and then to look at the tilting. and eh.. then to build the controller. because we had the fire on last night, there was still some hot in the cylinder this morning, and for most of the day. what should a controller do if the top of the cylinder is, 35° but the bottom is 20°, let's say. should the pump come on when the panel is at about 26, or slightly higher? i was playing with it today (there were points when the panel got fairly hot) and it seemed that when the panel was around the same as the top of the cylinder, it was almost seeming to cool the top of the cylinder by pumping through. i also wonder if the pump is too vigourous. i have it on the lowest setting, but i wonder if i'd get a better effect from the panel if the cold water was returning into it more slowly, allowing more time for mixing inside. might run the fire again tonight to check the new non return valve. would be nice to know that's working ok. thanks again, bye for this evening i think. here is today..
|
|
|
Post by cye on Aug 5, 2014 15:43:24 GMT -5
looking very neat now steve!
tec7 eh, a man after paddy bloomer's heart i see. great stuff, and virtually the same stuff (MS polymer) as evostick 'nail & seal', though the latter only available in white. overpaintable and sticks to everything , wet or dry, except pvc as far as i can tell.
if the panel and the pipework is well insulated, then in theory too fast a pump will not cause a problem but it will be a waste of energy. though it may cool the panel quicker than a pump controller can respond. with poor lagging and panel insulation too fast a pump will create a net loss of heat from your tank. a pump should come on when the top of the panel (top right hand side in your case) is 5 or 6 c above the temp of the tank at the level equal to or just above the lowest turn in the solar coil (which in your case is also the dual-purpose boiler coil), but don't measure the coil temp itself, measure the tank wall temp, & as far around the tank as you can get from the coil port. if you try and run the pump at a greater temperature difference, then this runs the panel hot, which results in a less efficient collector panel, particularly an issue for a flat plate collector such as your's. all panels are more efficient at capturing solar energy the closer they are to ambient air temp.
i guess we shouldn't read too much into the performance until, as you say, you've lagged everything up and got a proper controller in place. if you PM me your address, the co-op has a spare, albeit used but functional, 6w 12v pump it will donate to the project. the central heating pump is overkill and wasteful of electricity, probably 35w at the lowest setting?
rgds
cye
|
|
|
Post by cravings on Aug 6, 2014 4:25:37 GMT -5
ugh.
really thought i had got past some stuff last night... but today, the warm water in the panel doesn't seem to be pumping around. i can only presume it's the non return valve. i'm certain i put it on the right way... but the warm water definitely isn't moving. i've checked all the isolation valves are open, i've turned the pump up to its highest setting, but the water definitely isn't moving. the whole system is definitely full, i'm sure of that, water bleeds out of the top of the panel as expected when i open the valve.
and i can feel the pump vibrating, i can feel the pipes going to / from the back boiler vibrating slightly.. i'm a bit confused now. i really didn't want to have to drain and fill it again.
|
|
|
Post by cravings on Aug 6, 2014 7:23:11 GMT -5
ok phew, got it moving. closed the panel, drained the rest, had a look at the nrv, it seemed fine.. filled again. pump still not pumping... water in the panel got to above 70°.. disconnected one join on the hot out to see what happened... after a few seconds delay, finally hot water came spurting out everywhere, burnt my finger slightly.. managed to get it back together.
not really sure what happened.. maybe an airlock somewhere.. but it seems to be cleared now. obviously wasted a lot of the hottest water, but i'm glad i have it circulating again, pump is on the lowest setting and the water in the panel is going up very slowly again, and the water in the cylinder is going up too.
phew.
|
|
|
Post by cravings on Aug 7, 2014 15:23:40 GMT -5
hmm. today went fairly well, solar wise. however, have a fire lit now to top it up, and am having the same thing with the new nrv. they don't seem to stop the hot water from flowing up to the panel. again, i've closed the ball / isolator valve immediately after the nrv, and this has worked fine.
need to look at this. the other thing i've noticed is that the temperature of the water in the back boiler has been going up when the solar is pumping.. had thought i could put a non return valve against this flow.. but now i don't know if that's the answer.
anyway, i carry on.
|
|
|
Post by cravings on Aug 9, 2014 7:53:49 GMT -5
so yeah. the lagging is working. it definitely does heat the water a bit every day. the issue with the non return valve not working is puzzling. i talked to paddy about it.. he doesn't know.
also, the water in the back boiler is heating with the solar panel.. so a lot of the hot water from that loop is seemingly flowing through the back boiler too. so i'm losing a lot of heat there too.
i did think about putting another non return valve to stop hot water flowing into the top of the back boiler, but i think i'd definitely need to run the existing heating circulation pump to heat water by the back boiler then. at the moment it's nice that you light a fire, and no pump is required to heat the cylinder (except for the issue of the hot water flowing up to the panel.
all in all, it does seem like sharing the boiler coil is fairly troublesome... so a next step might be to look at fitting a dual coil cylinder. i do have one lying around, but it has a leak.. i'll look at it more carefully soon enough.
a question though: is there a reason why i couldn't have 2 separate heating systems (back boiler and solar) with 2 separate cylinder coils, but sharing a feed / expansion tank at the top? would that cause any issues?
|
|
|
Post by cye on Aug 10, 2014 8:54:18 GMT -5
so yeah. the lagging is working. it definitely does heat the water a bit every day. the issue with the non return valve not working is puzzling. i talked to paddy about it.. he doesn't know. also, the water in the back boiler is heating with the solar panel.. so a lot of the hot water from that loop is seemingly flowing through the back boiler too. so i'm losing a lot of heat there too. i did think about putting another non return valve to stop hot water flowing into the top of the back boiler, but i think i'd definitely need to run the existing heating circulation pump to heat water by the back boiler then. at the moment it's nice that you light a fire, and no pump is required to heat the cylinder (except for the issue of the hot water flowing up to the panel. all in all, it does seem like sharing the boiler coil is fairly troublesome... so a next step might be to look at fitting a dual coil cylinder. i do have one lying around, but it has a leak.. i'll look at it more carefully soon enough. a question though: is there a reason why i couldn't have 2 separate heating systems (back boiler and solar) with 2 separate cylinder coils, but sharing a feed / expansion tank at the top? would that cause any issues? Some NRVs might need a minimum pressure to close/seat fully. perhaps google the make/model of what you're using to ascertain its rating? regarding the loss of solar into the back boiler, you may consider trying a flap (aka tilt) type NRV in the boiler loop. This kind of NRV works with 'gravity fed' & is allowable under regulations in RoI, and may enable you to use your back boiler without the pump. You are the first person I've encountered who's attempted to dual-purpose the boiler coil, so we have no experience of it on this forum. An external plate heat exchanger would bypass all these problems and would be less hassle than swapping tanks or using a second tank. I see no reason why you could not pressurise two separate coils from the same F&E tank, apart from the unlikely event that if they are both hot at the same time then the expansion tank may overflow as it will have been sized for just the boiler loop's volume of water. regards cye
|
|
|
Post by cravings on Aug 12, 2014 7:33:44 GMT -5
the NRV is a sprung plunger type, which is normally closed. so it takes forward pressure from the solar flow side to push it open, the strange thing is that the flow of hot is in the other direction.
anyway, i'm putting aside all these issues for the moment and want to concentrate on getting the pump controller done.
so as you've said, the pump should run when the panel temp is 6 degrees hotter than the lowest turn on the cylinder coil. grand. but.. what if the top of the tank is hotter?
today, for example, the top of the cylinder is about 45, and the bottom is about 33. so if the pump runs when the panel is 39.. will this not have the effect of somewhat (slightly?) cooling the top of the cylinder?
my dad's house, he has a pro solar water install, and his pumps at 5 degrees, but his twin coil tank has the solar coil right at the bottom, so i presume in his case it pumps when the bottom of the cylinder is 5 degrees under the roof.
|
|
|
Post by cye on Aug 12, 2014 11:22:19 GMT -5
the NRV is a sprung plunger type, which is normally closed. so it takes forward pressure from the solar flow side to push it open, the strange thing is that the flow of hot is in the other direction. anyway, i'm putting aside all these issues for the moment and want to concentrate on getting the pump controller done. so as you've said, the pump should run when the panel temp is 6 degrees hotter than the lowest turn on the cylinder coil. grand. but.. what if the top of the tank is hotter? today, for example, the top of the cylinder is about 45, and the bottom is about 33. so if the pump runs when the panel is 39.. will this not have the effect of somewhat (slightly?) cooling the top of the cylinder? my dad's house, he has a pro solar water install, and his pumps at 5 degrees, but his twin coil tank has the solar coil right at the bottom, so i presume in his case it pumps when the bottom of the cylinder is 5 degrees under the roof. delta T - re you dad's house (twin coil) yes what you say is correct. re your setup, if you could instead measure the tank temp at the top turn of the boiler coil this would probably avoid the scenario you describe, but then you will (on average) be running your panel at hotter temperatures which is less efficient (for collector efficiency, particularly your type of collector). tank top temp is less important than the temp at the top of coil as the hot water layer will be quite distinct (stratification). a fair compromise could be to use the middle of the boiler coil. in the scenario where a tank is very well insulated, gets heat from nowhere but solar, and has little or no hot water usage during the best daylight hours, with the bulk of hot water used by early morning, then the bottom turn of the coil is the best place for the tank sensor from an efficiency perspective. additional electronic control could perhaps help too - e.g., a second differential controller, say working only when the pump has been on for X seconds (time delay from pump on), and detecting the difference between coil input and output temps, could be used to guard against the undesired cooling scenario you describe. NRV - your NRV is clearly not normally closed as it is letting flow through in the wrong direction, though it may seem that it is normally closed when you blow through it. you are perhaps fulfilling some minimum pressure requirement to seat the plunger when you blow, or it could be dodgy stock (if new), or wear (if it's not new), it certainly should not be doing what it's doing.
|
|
|
Post by cye on Aug 13, 2014 2:05:09 GMT -5
hi steve, now you've got me thinking about that tall boiler coil of your's.
if you are handy enough to be making your own REUK style pump differential temp controller, then why not make two:
[1] one to control the pump, connected to one sensor for the panel, and two sensors at the boiler coil, though only only tank sensor connected to the pump controller at any given time.
[2] the switching between the two tank sensors mentioned above would be via a changeover relay (DPDT) controlled by your second differential temp controller. The latter would monitor two sensors, one at the base of the coil and one at the top of the coil. when the top-of-coil temp is say X degrees hotter than the base, then it tells the pump controller to use the top sensor, and it does this by flipping the changeover relay.
For such a setup you will need 5 temp sensors, 3 for the pump controller and 2 for the controller which determines which tank sensor the first controller uses. Plus your dual pole dual throw relay to do the change-over between the pump controller's sensors.
Just an idea!
|
|